Fabrication and Characterisation of Calcium Sulphate Hemihydrate Enhanced with Zn- or B-Doped Hydroxyapatite Nanoparticles for Hard Tissue Restoration

A composite based on calcium sulphate hemihydrate enhanced with Zn- or B-doped hydroxyapatite nanoparticles was fabricated and evaluated for bone graft applications. The investigations of their structural and morphological properties were performed by X-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy, scanning electron microscopy (SEM), and energy dispersive X-ray (EDX) spectroscopy techniques. To study the bioactive properties of the obtained composites, soaking tests in simulated body fluid (SBF) were performed. The results showed that the addition of 2% Zn results in an increase of 2.27% in crystallinity, while the addition of boron causes an increase of 5.61% compared to the undoped HAp sample. The crystallite size was found to be 10.69 ± 1.59 nm for HAp@B, and in the case of HAp@Zn, the size reaches 16.63 ± 1.83 nm, compared to HAp, whose crystallite size value was 19.44 ± 3.13 nm. The mechanical resistance of the samples doped with zinc was the highest and decreased by about 6% after immersion in SBF. Mixing HAp nanoparticles with gypsum improved cell viability compared to HAp for all concentrations (except for 200 µg/mL). Cell density decreased with increasing nanoparticle concentration, compared to gypsum, where the cell density was not significantly affected. The degree of cellular differentiation of osteoblast-type cells was more accentuated in the case of samples treated with G+HAp@B nanoparticles compared to HAp@B. Cell viability in these samples decreased inversely proportionally to the concentration of administered nanoparticles. From the point of view of cell density, this confirmed the quantitative data.


Introduction
Bones are the major part of the musculoskeletal system that supports body weight, performs motion, and protects the internal organs [1]. They have functional adaptation ability, meaning that bones can adjust their mass and architecture according to their mechanical environment or other mechanical stimuli. Bone grafting is one of the most commonly used strategies for treating bone defects and is widely applied for bone regeneration in orthopaedic surgeries. An ideal bone graft must augment the process of bone healing Since the selective substitution of its cations (Ca 2+ ) and anions (OH − and/or PO 4 3− ) is allowed in HAp, it can become an ideal host for ionic doping [23]. Because HAp is weak in osteoinduction, ion doping can improve its biological activities. Different dopants have been studied, and research is ongoing to tailor synthetic HAp for different medical applications [24]. Zinc is an essential mineral that is also the second-most abundant mineral in the human body. Zinc deficiency can affect the central nervous, skeletal, and reproductive systems, as well as physical growth, and increase the risk of infection. Hydroxyapatite doped with zinc has been extensively studied in the last few years. Uysal et al. [24] published an extensive review of the different synthesis methods and sintering parameters for doped hydroxyapatite published during the last 20 years [18,[25][26][27][28][29][30]. Negrila et al. [25] and Predoi et al. [26] used sol-gel to produce HAp doped with Zn 2+ . Predoi et al. [26] also analysed the influence of the stability of Zn-HAp solutions on antibacterial properties. They found that Zn content has a significant impact on solution stability and prevents bacterial colonisation. Begam et al. [27] found that Zn 2+ doping changes the lattice parameter of HAp, increasing cell adhesion and growth.
Boron is an important trace element for plants, but it is not as important for animals. However, boron has been shown to play a role in osteoblastic driving [31]. Tunçay et al. [32] first attempted microwave-assisted biomimetic precipitation of B-HAp and showed that B-HAp accelerates cell attachment and differentiation and facilitates early mineralisation.
The present study is focused on the synthesis of composites prepared by mixing CSH with Zn-or B-doped hydroxyapatite and the investigation of their structural and morphological properties by X-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy, scanning electron microscopy (SEM), and energy dispersive X-ray (EDX) spectroscopy techniques. The results provided from this study could lead to the production of bone cement that could be used as complementary materials for filling bone defects and for their effective healing. To study the bioactive properties of the obtained composites, soaking tests in simulated body fluid (SBF) were performed. Moreover, the biological behaviour and cell proliferation of osteoblast-like MG-63 cells were investigated following incubation with nanostructured CSH-zinc/boron-doped hydroxyapatite composites.
After mixing, the pH of the solution was measured (Table 1) and adjusted to 10.5 with the addition of ammonium hydroxide (NH 4 OH). Afterward, the mixture was subjected to a hydrothermal treatment for 12 min at 120 • C, washed with distilled water, and dried at 60 • C for 24 h. To obtain semi-hydrate gypsum powder, calcium sulphate dihydrate was subjected to hydrothermal treatment at 132 • C for 30 min and left overnight at 50 • C in the oven. The composites were obtained by mixing calcium sulphate with the previously obtained hydroxyapatites. The composition and notations of the obtained samples are presented in Table 2. The composite was obtained by mixing undoped or doped hydroxyapatite with gypsum (1:1 weight ratio) and a 10% glycerol aqueous solution. The amount of solution used for each mixture was dosed judiciously to keep the same paste consistency.
The necessary volume of liquid was related to the size of the powders obtained. The doped hydroxyapatite showed smaller particles. The obtained mixture was poured into a cylindrical mould (ø = 10 mm and h = 10 mm) and characterised after hardening.

Samples Characterisation
The X-ray diffraction (XRD) technique was used to determine the degree of crystallinity, the crystallite size, and the phases present in the samples. The analysis was carried out using a PANalytical Empyrean diffractometer (Almelo, Netherlands) at room temperature with a characteristic Cu X-ray tube (λCuKα1= 1.541874 Å). The samples were scanned in a Bragg-Brentano geometry with a scan step increment of 0.02 • and a counting time of 100 s/step. The XRD patterns were recorded in the 2θ angle range of 5-80 • . Rietveld quantitative phase analysis was performed using the X Pert High Score Plus 3.0 software (PANalytical, Almelo, The Netherlands). After refining, values were obtained between 1.44% and 1.78% for goodness of fit, 6.65% and 7.08% for Rexpected, and 6.67% and 7.59% for Rprofile. The crystallite size was determined by Debye-Scherer Equation (1) where s = crystallite size (nm), k = the Scherrer constant (0.98), λ = denotes the wavelength (0.154 nm), and β = the full width at half maximum (FWHM) in radians Morphological aspects were studied via scanning electron microscopy (SEM) with a Quanta Inspect F50 microscope coupled with an energy dispersive spectrometer (EDS) and a Titan Themis 200 transmission electron microscope (TEM) with a line resolution of 90 pm in high-resolution transmission electron microscopy (HRTEM) mode (Thermo Fisher, Eindhoven, The Netherlands). The mechanical compression strength was determined using the Shimadzu Autograph AGS-X 20kN (Shimadzu, Tokyo, Japan) equipment. Fourier transform infrared spectroscopy (FTIR) investigation was performed using a Nicolet iS50R spectrometer (Thermo Fisher, Waltham, MA, USA). The measurements were made at room temperature using the total reflection attenuation module. Each sample was scanned 32 times between 4000 and 400 cm −1 , at a resolution of 4 cm −1 .
In order to evaluate the stability in a wet environment [33], the cylinders obtained from each composite were placed in 20 mL of simulated body fluid with pH = 7.4 (SBF) prepared according to the Kokubo recipe [34]. The initial mass was recorded, and the samples were immersed in SBF solution. After a certain amount of immersion, samples were removed, dried, and weighed again. The mass loss could be calculated using Equation (2): where w i = sample weight before immersion and w t = dried sample weight after t min of immersion in SBF. The mechanical compressive strength was determined by pressing the samples until the breaking point with a speed of 1 mm/min using Shimadzu Autograph AGS-X 20kN equipment (Shimadzu, Tokyo, Japan). The test was made in triplicate on samples hardened for 3, 7, and 28 days, according to the standard [35]. In addition to the initial composites (hardened in the air), samples immersed for 72 h in SBF were also tested.
The hydroxyapatite nanoparticles were suspended in deionised water at a concentration of 0.012 g/mL by dispersing with an ultrasound probe and then sterilised by gamma radiation. MG-63 cells were seeded in 96-well plates at a concentration of 2000 cells/well and incubated under standard conditions to allow cell attachment. After 24 h, the culture medium was removed and replaced with culture medium with nanoparticles at different concentrations (0, 25, 50, 100, and 200 µg/mL, previously prepared by dispersing in complete culture medium with an ultrasound bath). The cells were then incubated in the presence of hydroxyapatite nanoparticles for 7 days under standard temperature and humidity conditions. Following incubation in the presence of nanoparticles, investigations related to cell morphology, viability, and cell differentiation were carried out. Investigations of cell viability and proliferation were performed using the MTT tetrazolium salts assay (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) [36]. Seven days after treatment, the nanoparticle solution was removed and replaced with 10% MTT (5 mg/mL) in cell culture medium. After 2 h incubation, the reacted formazan crystals were solubilised using DMSO, and the absorbance of the supernatant was measured at 570 nm.
Cell differentiation was tested using Alizarin red assay [37]. Following incubation, the nanoparticle-containing cell culture medium was removed, and cells were fixed using 4% paraformaldehyde in PBS for 1 h. After this, the supernatant was removed, and cells were washed with PBS several times to remove any residues of nanoparticles left. Following this step, 40 mM Alizarin red was added to each well and incubated for 45 min at room temperature. Then, the supernatant was removed, and cells were thoroughly washed with deionised water. This method is based on the specific staining of Ca deposits in the cells resulting from osteoblast cell differentiation. The stained Ca deposits were then dissolved in 10% acetic acid, and the supernatant was collected and heated at 85 • C for 10 min. Following centrifugation at 20,000× g for 10 min, the supernatant was neutralised with 10% ammonium hydroxide. The absorbance was measured at 405 nm.
Experiments were performed in triplicate, and cell viability and differentiation ability, respectively, were calculated by relating the data obtained for each sample to the negative control samples (for which a value of 100% was assigned). Data were expressed as ± SEM (standard error of the mean), and their statistical evaluation was performed using the Student t-test function, for which * p < 0.05, ** p < 0.01, and *** p < 0.001. There were blank samples (nanoparticles without cells at the investigated concentrations) for all quantitative determinations, whose absorbance was subtracted from that of cellular samples.

HAp, HAp@Zn and HAp@B Powder Characterisation
The XRD profiles of HAp, HAp@B, and HAp@Zn prepared powders are displayed in Figure 1. The XRD pattern of HAp shows a single crystalline phase identified through the characteristic diffraction maxima. The phase identification is performed using the PDF 04-008-4763 file. Miller indices corresponding to the position of the crystallographic planes and directions could be associated with each important peak, such as (002), (121), (030), (222), and (123). The X-ray diffractogram indicates the presence of two high-intensity peaks located near 26 • and 32 • , which are in good agreement with the literature data for HAp [38,39].

HAp, HAp@Zn and HAp@B Powder Characterisation
The XRD profiles of HAp, HAp@B, and HAp@Zn prepared powders are displayed in Figure 1. The XRD pattern of HAp shows a single crystalline phase identified through the characteristic diffraction maxima. The phase identification is performed using the PDF 04-008-4763 file. Miller indices corresponding to the position of the crystallographic planes and directions could be associated with each important peak, such as (002), (121), (030), (222), and (123). The X-ray diffractogram indicates the presence of two high-intensity peaks located near 26° and 32°, which are in good agreement with the literature data for HAp [38,39].  Figure 1b shows the magnification of the area with the most intense peaks for the studied samples. The diffractograms for the HAp@Zn and HAp@B samples are similar to those for the HAp sample. However, the diffraction peaks are smaller in intensity, with the maximum being shifted and broader, which, according to Miyaji et al. [40], suggests that the crystallinity of apatite increases in the presence of a dopant.
In order to determine the degree of crystallinity, the average crystallite size, the structural microstrains, and the volume of the elementary cell, the Rietveld analysis was performed. The results of the refining are presented in Table 3. As can be seen, following the Rietveld analysis, the addition of the dopant tends to determine an increase in the crystallinity of powders. The addition of 2% Zn increases  Figure 1b shows the magnification of the area with the most intense peaks for the studied samples. The diffractograms for the HAp@Zn and HAp@B samples are similar to those for the HAp sample. However, the diffraction peaks are smaller in intensity, with the maximum being shifted and broader, which, according to Miyaji et al. [40], suggests that the crystallinity of apatite increases in the presence of a dopant.
In order to determine the degree of crystallinity, the average crystallite size, the structural microstrains, and the volume of the elementary cell, the Rietveld analysis was performed. The results of the refining are presented in Table 3. As can be seen, following the Rietveld analysis, the addition of the dopant tends to determine an increase in the crystallinity of powders. The addition of 2% Zn increases crystallinity by approximately 2.27%, while the addition of B causes an increase of approximately 5.61% compared to the undoped HAp sample. Also, the crystallite size decreases as a result of doping. In the case of B, a crystallite size of 10.69 ± 1.59 nm is recorded, and in the case of Zn doping, the size value reaches 16.63 ± 1.83 nm, while undoped HAp has a crystallite size value of 19.44 ± 3.13 nm.
Considering the atomic radius OF the dopants (Zn = 135 pm and B = 85 pm), their addition to the crystalline network of hydroxyapatite determines the change in the volume of the elemental cell and the microstrains of the structure. Since B has an atomic radius closer to that of P (100 pm), it is assumed to prefer its positions in the network and not the Ca (180 pm) positions. Therefore, a stronger tension caused by crystal defects present in the network is observed, as proven by the microstrain value of 0.86% compared to undoped HAp, which shows a value of 0.47%.
In order to complete the information from XRD studies related to the structure of the samples, Raman spectroscopy investigations were performed. The obtained spectra are presented in Figure 2. Considering the atomic radius OF the dopants (Zn = 135 pm and B = 85 pm), their addition to the crystalline network of hydroxyapatite determines the change in the volume of the elemental cell and the microstrains of the structure. Since B has an atomic radius closer to that of P (100 pm), it is assumed to prefer its positions in the network and not the Ca (180 pm) positions. Therefore, a stronger tension caused by crystal defects present in the network is observed, as proven by the microstrain value of 0.86% compared to undoped HAp, which shows a value of 0.47%.
In order to complete the information from XRD studies related to the structure of the samples, Raman spectroscopy investigations were performed. The obtained spectra are presented in Figure 2. Following Raman spectroscopy, it is possible to observe the bands located at 431 and 445 cm −1 , which can be attributed to the bending vibration of the O-P-O bonds doubly degenerated (ν2). The bands located at values of 578 and 590 cm −1 were attributed to the bending vibration of the O-P-O bonds, which triply degenerated (ν4). The intense band located at 961 cm −1 was attributed to the stretching of the non-degenerate P-O symmetric bond (ν1). Additionally, the bands located at values of 1046 and 1075 cm −1 were attributed to the antisymmetric stretching of the P-O bond triple degenerated (ν3) [41,42].
Analysing the band located at 961 cm −1 , characteristic of the vibration of PO4 tetrahedra, it is observed that it widens with the addition of the dopant. The ratio between the intensities of the ν1 and ν2 vibration modes decreases from 3.59 for HAp to 2.98 for HAp@Zn, respectively, to 2.48 for HAp@B, which indicates the distortion and increase in the disorder degree of the crystalline network [43].
The shape and size of the particle aggregates were determined by SEM. The images obtained for HAp, HAp@Zn, and HAp@B are presented in Figure 3 at different magnifications. Following Raman spectroscopy, it is possible to observe the bands located at 431 and 445 cm −1 , which can be attributed to the bending vibration of the O-P-O bonds doubly degenerated (ν2). The bands located at values of 578 and 590 cm −1 were attributed to the bending vibration of the O-P-O bonds, which triply degenerated (ν4). The intense band located at 961 cm −1 was attributed to the stretching of the non-degenerate P-O symmetric bond (ν1). Additionally, the bands located at values of 1046 and 1075 cm −1 were attributed to the antisymmetric stretching of the P-O bond triple degenerated (ν3) [41,42].
Analysing the band located at 961 cm −1 , characteristic of the vibration of PO 4 tetrahedra, it is observed that it widens with the addition of the dopant. The ratio between the intensities of the ν1 and ν2 vibration modes decreases from 3.59 for HAp to 2.98 for HAp@Zn, respectively, to 2.48 for HAp@B, which indicates the distortion and increase in the disorder degree of the crystalline network [43].
The shape and size of the particle aggregates were determined by SEM. The images obtained for HAp, HAp@Zn, and HAp@B are presented in Figure 3 at different magnifications. In the case of the HAp, medium-sized particles are observed around the val 67.11 nm, while the values decrease with the addition of dopants, reaching 54.67 nm the sample doped with B and 52.64 nm for the sample doped with Zn. Also, due t small size and the pseudo-acicular shape, the samples show an accentuated agglom tion tendency.
The presence of characteristic hydroxyapatite elements (Ca, P) and dopants demonstrated by the EDS spectra and the elemental composition of the powder prese in Table 4. Boron was found in HAp particles in a proportion of 3.72 wt.% (Table 4), whi was found in a proportion of 0.33 wt.%. The presence of carbon can be attributed to sa preparation for SEM analysis. However, it can also be attributed to the possible carb tion of the HAp particles under the influence of the synthesis conditions and the CO sent in the atmosphere [25]. In the case of the HAp, medium-sized particles are observed around the value of 67.11 nm, while the values decrease with the addition of dopants, reaching 54.67 nm for the sample doped with B and 52.64 nm for the sample doped with Zn. Also, due to the small size and the pseudo-acicular shape, the samples show an accentuated agglomeration tendency.
The presence of characteristic hydroxyapatite elements (Ca, P) and dopants was demonstrated by the EDS spectra and the elemental composition of the powder presented in Table 4. Boron was found in HAp particles in a proportion of 3.72 wt.% (Table 4), while Zn was found in a proportion of 0.33 wt.%. The presence of carbon can be attributed to sample preparation for SEM analysis. However, it can also be attributed to the possible carbonation of the HAp particles under the influence of the synthesis conditions and the CO 2 present in the atmosphere [25].
The TEM images obtained on the HAp samples highlight well-defined particles with a polyhedral, irregular shape, with sizes between 30-70 nm for the HAp (Figure 4a), 15-75 nm for the HAp@Zn sample, and 40-65 nm for the HAp@B sample (Figure 4d). Also, in the case of doping with Zn, the particles are presented in acicular hexagonal form (Figure 4g). The polycrystalline character of all samples is highlighted by SAED analysis (Figure 4c,f,i). The TEM images obtained on the HAp samples highlight well-defined particles w a polyhedral, irregular shape, with sizes between 30-70 nm for the HAp (Figure 4a), 75 nm for the HAp@Zn sample, and 40-65 nm for the HAp@B sample (Figure 4d). A in the case of doping with Zn, the particles are presented in acicular hexagonal form ( ure 4g). The polycrystalline character of all samples is highlighted by SAED analysis ( ure 4c,f,i).
Performing the measurements on the HRTEM images (Figure 4b,e,h), it was fou that the distances between the atomic chains vary between 3.5 Å for the HAp sample Å for HAp@Zn, and 3.3 Å for HAp@B.  Performing the measurements on the HRTEM images (Figure 4b,e,h), it was found that the distances between the atomic chains vary between 3.5 Å for the HAp sample, 3.4 Å for HAp@Zn, and 3.3 Å for HAp@B.
The mapping of the main elements on the sample surface is presented in Figure 5. As shown in Figure 5, in addition to the characteristic elements of hydroxyapatite (Ca, P, and O), the presence of carbon could also be identified in the sample. This can be attributed to superficial carbonation of the sample, most likely during synthesis following interaction with atmospheric CO 2 . The elemental distribution of the dopants indicates that they are distributed uniformly among the powder particles.

Gypsum Analysis
Gypsum is a biomaterial that has been used as bone cement for many years [44]. It is a versatile material used in several medical applications due to its biocompatibility, biodegradability, and easy availability. Gypsum is a naturally occurring mineral made up of calcium sulfate. It is widely used in construction and medical applications, such as making casts and moulds. Gypsum can also be easily modified to adjust its properties, which makes it an excellent candidate for use in bone cements that must meet specific mechanical and biological requirements. When used as bone cement, gypsum is mixed with water to create a paste-like substance that can be moulded to the shape of the bone. The paste then hardens to create a rigid bond, much like traditional cement. Gypsum can also be reinforced with other materials, such as carbon fibers, to increase its strength and durability. Figure 6 shows the X-ray diffractogram of the gypsum powder. The specific peaks of calcium sulphate hemihydrate are observed, with the most prominent peak at an angle of 11.7 • , according to the PDF 04-015-7420 file. Also, each important peak has Miller indices associated with it, corresponding to the plane position and crystallographic direction associated with the peak.

Gypsum Analysis
Gypsum is a biomaterial that has been used as bone cement for many years [44]. It is a versatile material used in several medical applications due to its biocompatibility, biodegradability, and easy availability. Gypsum is a naturally occurring mineral made up of calcium sulfate. It is widely used in construction and medical applications, such as making casts and moulds. Gypsum can also be easily modified to adjust its properties, which makes it an excellent candidate for use in bone cements that must meet specific mechanical and biological requirements. When used as bone cement, gypsum is mixed with water to create a paste-like substance that can be moulded to the shape of the bone. The paste then hardens to create a rigid bond, much like traditional cement. Gypsum can also be reinforced with other materials, such as carbon fibers, to increase its strength and durability. Figure 6 shows the X-ray diffractogram of the gypsum powder. The specific peaks of calcium sulphate hemihydrate are observed, with the most prominent peak at an angle of 11.7°, according to the PDF 04-015-7420 file. Also, each important peak has Miller indices associated with it, corresponding to the plane position and crystallographic direction associated with the peak. After carrying out the FTIR analysis, the specific graph indicates the structural composition of calcium sulphate. Thus, one can observe the vibration bands characteristic of stretching the O-H bond at values 3200-3600 cm −1 , but also between 1500-1700 cm −1 , and the sulphate bond (S-O) characteristic bands of between 850-1230 and also 400-800 cm −1 [45,46].
The EDS analysis performed on the plaster sample shows specific chemical elements such as oxygen (O) at 0.53 keV, sulphur (S) at 2.31 keV, and calcium (Ca) at a value of 3.7 keV. The percentage composition of the elements is presented in Figure 7. After carrying out the FTIR analysis, the specific graph indicates the structural composition of calcium sulphate. Thus, one can observe the vibration bands characteristic of stretching the O-H bond at values 3200-3600 cm −1 , but also between 1500-1700 cm −1 , and the sulphate bond (S-O) characteristic bands of between 850-1230 and also 400-800 cm −1 [45,46].
The EDS analysis performed on the plaster sample shows specific chemical elements such as oxygen (O) at 0.53 keV, sulphur (S) at 2.31 keV, and calcium (Ca) at a value of 3.7 keV. The percentage composition of the elements is presented in Figure 7.

Composites Material Characterisation
After hardening, the composites, obtained by mixing the gypsum with the HAp-type particles and the aqueous solution, were matured for 28 days and characterised by FTIR. The results are presented in Figure 8.

Composites Material Characterisation
After hardening, the composites, obtained by mixing the gypsum with the HAp-type particles and the aqueous solution, were matured for 28 days and characterised by FTIR. The results are presented in Figure 8.

Composites Material Characterisation
After hardening, the composites, obtained by mixing the gypsum with the HAp-type particles and the aqueous solution, were matured for 28 days and characterised by FTIR. The results are presented in Figure 8.  Scanning electron microscopy was performed on each composite sample, and the results are shown in Figure 9.
Nanomaterials 2023, 13, x FOR PEER REVIEW 13 and between 1370-1520 cm −1 , and the PO4 −3 group between 900-1220 cm −1 and bet 400-740 cm −1 . Scanning electron microscopy was performed on each composite sample, and t sults are shown in Figure 9. The morphology of the composite samples is similar and is predominantly fo by the characteristic forms of calcium sulphate dihydrate. Samples are formed by cr with sizes between 200-350 nm. As shown in Figure 9, hydroxyapatite does not si cantly influence the behaviour of the samples; HAp particles are uniformly distribut the surface of the gypsum crystals as a result of the mechanical homogenisation pr At low magnifications, one can see the porosity obtained by the intercalation of thes yhedral crystals.
To perform in vitro tests, the composites were moulded into cylindrical shape after hardening, immersed in SBF solution. The weight loss variation of composite ples G+HAp, G+HAp@B, and G+HAp@Zn during 72 h of immersion in SBF is pres in Figure 10. The morphology of the composite samples is similar and is predominantly formed by the characteristic forms of calcium sulphate dihydrate. Samples are formed by crystals with sizes between 200-350 nm. As shown in Figure 9, hydroxyapatite does not significantly influence the behaviour of the samples; HAp particles are uniformly distributed on the surface of the gypsum crystals as a result of the mechanical homogenisation process. At low magnifications, one can see the porosity obtained by the intercalation of these polyhedral crystals.
To perform in vitro tests, the composites were moulded into cylindrical shapes and, after hardening, immersed in SBF solution. The weight loss variation of composite samples G+HAp, G+HAp@B, and G+HAp@Zn during 72 h of immersion in SBF is presented in Figure 10.
The rate of degradation of the composites in the SBF environment is accelerated in the first 15 h when a decrease in the weight of the samples of approximately 37% is recorded. After 15 h, a decrease in degradation rate is observed, with the sample reaching a maximum of approximately 47% after 72 h of contact with the liquid. This rate of disintegration of the resulting paste may coincide with the beginning of cell proliferation to form new bone tissue.
It is also noted that G+HAp@Zn samples tend to disintegrate more slowly in the presence of SBF than the other types of studied composites that showed similar behaviour. The visual aspects of the samples before and after immersion in SBF for 72 h are shown in Figure 11, and SEM images are presented in Figure 12. It is also noted that G+HAp@Zn samples tend to disintegrate more slowly in th ence of SBF than the other types of studied composites that showed similar beh The visual aspects of the samples before and after immersion in SBF for 72 h are sh Figure 11, and SEM images are presented in Figure 12.   The rate of degradation of the composites in the SBF environment is accelerated in the first 15 h when a decrease in the weight of the samples of approximately 37% is recorded. After 15 h, a decrease in degradation rate is observed, with the sample reaching a maximum of approximately 47% after 72 h of contact with the liquid. This rate of disintegration of the resulting paste may coincide with the beginning of cell proliferation to form new bone tissue.
It is also noted that G+HAp@Zn samples tend to disintegrate more slowly in the presence of SBF than the other types of studied composites that showed similar behaviour. The visual aspects of the samples before and after immersion in SBF for 72 h are shown in Figure 11, and SEM images are presented in Figure 12.  Keeping the samples in SBF leads to a rearrangement of the microstructure. As can be seen in the SEM images, the gypsum crystals, initially formed following the hardening process, decrease in size and flatten due to the dissolution process. This process is followed by the mineralisation of the crystal surface through the deposition of apatite phases that lead to a more compact structure. This formation of new apatite phases is also confirmed by the FTIR analysis ( Figure 13).
The bands detected at 566, 601, 962, 1039, and 1089 cm −1 belong to the phosphate phase [47,48]. The band corresponding to the hydroxyl group can be observed at 631 and 3550 cm −1 [49], and the band attributed to water molecules is observed around 3050-3550 cm −1 .
Also, specific bands for carbonate groups were detected at 872, 1421, and 1467 cm −1, and the band of calcium sulphate is present at a value of 3380 cm −1 [50][51][52].
Compared to the FTIR performed on the composites before immersion in SBF (see Figure 8), a decrease in the characteristic bands for calcium sulphate and water and an increase in absorption for the PO 4 3− bands are observed. This indicates the formation of a new apatite phase. Nanomaterials 2023, 13, x FOR PEER REVIEW 15 of 21 Keeping the samples in SBF leads to a rearrangement of the microstructure. As can be seen in the SEM images, the gypsum crystals, initially formed following the hardening process, decrease in size and flatten due to the dissolution process. This process is followed by the mineralisation of the crystal surface through the deposition of apatite phases that lead to a more compact structure. This formation of new apatite phases is also confirmed by the FTIR analysis ( Figure 13).   Keeping the samples in SBF leads to a rearrangement of the microstructure. As can be seen in the SEM images, the gypsum crystals, initially formed following the hardening process, decrease in size and flatten due to the dissolution process. This process is followed by the mineralisation of the crystal surface through the deposition of apatite phases that lead to a more compact structure. This formation of new apatite phases is also confirmed by the FTIR analysis ( Figure 13).  Regarding the mechanical resistance of the samples at various hardening intervals (Table 5), it was found that the most resistant was the sample doped with Zn. Also, the values for the samples immersed in SBF are 1.08 MPa for the G+HAp sample, 0.74 MPa for the G+HAp@B sample, and 1.93 MPa for the G+HAp@Zn sample. From these values, it can be seen that the samples gain strength for up to 28 days. After introducing samples in SBF, these resistances decrease by about 6%. The biological evaluation of the gypsum/hydroxyapatite-modified nanoparticles was conducted for MG-63 osteoblast-like cells in terms of proliferation measurements and the ability of the cells to differentiate and form bone tissue following mineralisation. Osteoblast proliferation in the presence of nanovehicles is an important parameter to follow in terms of studying the stimulation abilities of osteoblasts. In this regard, the MTT tetrazolium-salt viability assay was performed following 7 days of incubation in the presence of nanoparticles. This method measures mitochondrial metabolism. The ability to metabolise the substance is proportional to cell viability. The analysis performed 7 days following cell seeding can also measure osteoblast cell proliferation in the presence of the nanoparticles because this incubation time is longer than one cell cycle.
The results are comparatively shown in Figure 14, depending on the type of HAp. In the case of (G)/HAp, the cell metabolic activity decreased proportionally with the increase in nanoparticle concentration (p < 0.05 for 200 µg/mL). Gypsum alone proved to have a biocompatible behaviour at all concentrations, as the cell viability did not decrease below the 70% threshold for any of the investigated concentrations. With the addition of gypsum to the hydroxyapatite nanoparticles, cell viability was considerably improved compared to HAp alone. At all concentrations, unless the highest concentration was employed in the study (200 µg/mL), a reduction in cell metabolism was observed. For 200 µg/mL of G+HAp, a reduction in mitochondrial metabolism was observed compared to HAp. However, the difference noticed between these two samples was not statistically significant (G+HAp vs. HAp, NS). Overall, all samples proved biocompatible at concentrations below 100 µg/mL. A statistically significant proliferation of osteoblast-like cells was noticed in the case of G+HAp at 50 µg/mL (p < 0.01 compared to the negative control).
The modification of HAp with B did not show any significant changes in the behaviour of the cells in terms of viability and proliferation compared to (gypsum)/HAp. The cells' metabolic activity decreases with the increase in nanoparticle concentration. HAp@B showed a biocompatible behaviour at concentrations of 100 µg/mL and above, while 200 µg/mL induced a cytotoxic effect in osteoblasts following 7 days of incubation (p < 0.05 compared to the negative control).
In the case of HAp@Zn nanoparticles, the reduction in cells' metabolism was more pronounced at all of the investigated concentrations compared to HAp and HAp@B samples. This inhibitory effect is probably induced by the presence of Zn in the composition of the nanoparticles, which is a well-known effect determined by Zn 2+ ion dissolution [53,54].
The addition of gypsum showed an improved effect on osteoblast cell viability at 25 µg/mL (HAp@Zn vs. G+HAp@Zn, p < 0.01).
Differentiation of osteoblast cells involves the mineralisation of the extracellular matrix in order to form bone matrix, similar to the in vivo environment [55]. The Ca deposits were evidenced in the nanoparticle-treated osteoblasts using a specific reaction between alizarin red and the mineralised areas in the extracellular matrix of the cells (Figure 15). Using this test, the native ability of HAp alone to induce differentiation of osteoblast cells at 7 days of incubation was clearly evidenced. For all samples, blank samples were prepared at equivalent concentrations in order to remove the possible interferences induced by the presence of nanoparticles alone. Making a correlation with MTT assay data, when differentiation occurs in osteoblast cells, proliferation is usually inhibited, and a reduction in the cells' metabolism takes place [56,57]. Thus, the reduction in metabolic activity of MG-63 exposed to HAp nanoparticles is correlated with differentiation data, where a significant amount of Ca deposits was measured.
pronounced at all of the investigated concentrations compared to HAp and HAp@B samples. This inhibitory effect is probably induced by the presence of Zn in the composition of the nanoparticles, which is a well-known effect determined by Zn 2+ ion dissolution [53,54].
The addition of gypsum showed an improved effect on osteoblast cell viability at 25 µg/mL (HAp@Zn vs. G+HAp@Zn, p < 0.01). Differentiation of osteoblast cells involves the mineralisation of the extracellular matrix in order to form bone matrix, similar to the in vivo environment [55]. The Ca deposits were evidenced in the nanoparticle-treated osteoblasts using a specific reaction between alizarin red and the mineralised areas in the extracellular matrix of the cells (Figure 15). Using this test, the native ability of HAp alone to induce differentiation of osteoblast cells at 7 days of incubation was clearly evidenced. For all samples, blank samples were prepared at equivalent concentrations in order to remove the possible interferences induced by the presence of nanoparticles alone. Making a correlation with MTT assay data, when differentiation occurs in osteoblast cells, proliferation is usually inhibited, and a reduction in the cells' metabolism takes place [56,57]. Thus, the reduction in metabolic activity of MG-63 exposed to HAp nanoparticles is correlated with differentiation data, where a significant amount of Ca deposits was measured.
Although gypsum alone inhibited the differentiation of osteoblasts in terms of calcium deposition, this suppressing effect did not exceed 30% for all concentrations (p < 0.05). In consequence, this effect was also translated to cells exposed to gypsum-modified HAp, where a decrease in Ca deposit production was measured compared to control cells. However, this effect was statistically significant only for 200 µg/mL concentration.
The addition of B to the composition of HAp nanoparticles clearly improved the ability of osteoblast-like cells to differentiate and mineralise the extracellular matrix in vitro. The amount of Ca deposits was significantly higher compared to negative control samples A similar effect of differentiation improvement was measured in the cases of HAp@Zn and G+HAp@Zn compared to the negative control. However, the addition of gypsum to the composition of HAp@Zn samples did not increase the amount of resulting calcium deposits as compared to HAp@Zn alone. Similarly, the differentiation of osteoblasts exposed to HAp@Zn and G+HAp@Zn was correlated with a reduction in the cells' metabolism compared to the negative control.

Conclusions
The study presents the results of the preparation and characterisation of the composite obtained by mixing gypsum with Zn-or B-doped hydroxyapatite nanoparticles for hard tissue restoration. Although gypsum alone inhibited the differentiation of osteoblasts in terms of calcium deposition, this suppressing effect did not exceed 30% for all concentrations (p < 0.05). In consequence, this effect was also translated to cells exposed to gypsum-modified HAp, where a decrease in Ca deposit production was measured compared to control cells. However, this effect was statistically significant only for 200 µg/mL concentration.
The addition of B to the composition of HAp nanoparticles clearly improved the ability of osteoblast-like cells to differentiate and mineralise the extracellular matrix in vitro. The amount of Ca deposits was significantly higher compared to negative control samples at all concentrations. The high ability of HAp@B and G+HAp@B to induce mineralisation in MG-63 cell culture was correlated with their metabolism inhibition.
A similar effect of differentiation improvement was measured in the cases of HAp@Zn and G+HAp@Zn compared to the negative control. However, the addition of gypsum to the composition of HAp@Zn samples did not increase the amount of resulting calcium deposits as compared to HAp@Zn alone. Similarly, the differentiation of osteoblasts exposed to HAp@Zn and G+HAp@Zn was correlated with a reduction in the cells' metabolism compared to the negative control.

Conclusions
The study presents the results of the preparation and characterisation of the composite obtained by mixing gypsum with Zn-or B-doped hydroxyapatite nanoparticles for hard tissue restoration.
The experiments showed that the addition of the dopant led to better crystallisation. By adding 2% Zn, a 2.27% increase in crystallinity was observed, while the addition of B led to an increase of 5.61% compared to HAp. The crystallite size decreases as a result of doping. For B, a crystallite size of 10.69 ± 1.59 nm was measured, and in the case of Zn, the size value was 16.63 ± 1.83 nm, compared to HAp, where the crystallite size value was 19.44 ± 3.13 nm.
In the case of the HAp, medium-sized particles are observed around the value of 67.11 nm, while the values decrease with the addition of dopants, reaching 54.67 nm for the sample doped with B and 52.64 nm for the sample doped with Zn. Also, due to their small size and pseudo-acicular shape, the samples show an accentuated agglomeration tendency.
After hardening, the composites, obtained by mixing the gypsum with the HAp-type particles and the aqueous solution, were matured for 28 days and characterised by FTIR.
The morphology of the composite samples is similar and is predominantly formed by the characteristic forms of calcium sulphate dihydrate. Samples are formed by crystals with sizes between 200-350 nm. HAp particles are uniformly distributed on the surface of the gypsum crystals due to the mechanical homogenisation process.
The rate of degradation of the composites in the SBF environment is accelerated in the first 15 h when a decrease in the weight of the samples by approximately 37% is recorded. After a decrease in degradation rate is observed, the sample reaches a maximum of approximately 47% after 72 h of contact with the liquid. This rate of disintegration of the resulting paste may coincide with the beginning of cell proliferation to form new bone tissue.
It is also noted that G+HAp@Zn samples tend to disintegrate more slowly in the presence of SBF compared to the other types of studied composites that showed similar behaviour. After the introduction of samples in SBF, these resistances decreased by about 6%.
The ability of HAp, HAp@B, and HAp@Zn nanoparticles, respectively, to mineralise MG-63 osteoblast-like cells was clearly highlighted at all concentrations involved in the study. This effect is correlated with the reduction in the cells' metabolism following the differentiation process, which normally takes place when the mineralisation of bone cells occurs.